Your Secret Weapon for MBA Applications: Using AI the Smart Way
(AND ONE CRUCIAL WARNING!)
Let’s be real: applying to business school is a marathon. Between the GMAT, endless essays, and trying to figure out which school is the one, it’s easy to feel overwhelmed. Now, there’s a new player on the field: Artificial Intelligence. Tools like ChatGPT and Gemini are everywhere, and you’re probably wondering, “Can this make my life easier? Can it help me get in?”
The answer is a resounding yes… but with a huge asterisk.
The conversation has thankfully moved beyond if you should use AI to how you should use it – schools are increasingly updating their ethics policies to reflect the inevitable use of LLMs in the application process. Think of AI not as a magic wand to write your application for you, but as a powerful, strategic co-pilot on your journey (Microsoft better be paying me the big bucks for this strategic name drop). By understanding what it does best, and, more importantly, where it falls short—you can leverage generative AI to ease the challenge of creating an application that’s polished, powerful, and authentically you.
So, let’s dive into how you can make AI your secret weapon without letting it dull your unique shine.
The AI Sweet Spot: Your Personal Editor, Research Guru, and (sometimes) Copywriter
So where should you let AI shine? Think of it as your super-powered assistant for two key areas: getting the grammatical details right and digging up relevant school-facing facts about the MBA programs you’re interested in.
First up, rule-based tasks. This is AI’s bread and butter. For years, we’ve had tools like Grammarly, but today’s Large Language Models (LLMs) are in a different league. They are phenomenal for polishing your essays and resume from a grammar point of view. Got a comma splice you can’t spot? Worried your sentence structure is getting repetitive? Feed your text to an AI. It will meticulously comb through it, catching grammatical errors and suggesting improvements based on the established rules of language. It’s like having a world-class copy editor on call, 24/7.
The second sweet spot, and where things get really strategic, is using AI as your personal research assistant. We’ve all been there—spending hours clicking through a university’s website, trying to connect the dots between your goals and their specific offerings. This is where you can save a ton of time.
Information about a school—its stated values, its faculty’s research publications, its core curriculum—is a relatively stable set of data, making it perfect for an AI to analyze. You can ask LLMs incredibly specific questions like:
- “Which professors at Chicago Booth are researching the intersection of finance and technology?”
- “Based on recent news from the school’s website, what are the key initiatives of the Social Impact Club at Wharton?”
- “What are the most popular student-led conferences at London Business School?”
Answering these questions manually could take hours of detective work. An AI can synthesize that information in minutes. This aligns with Karpathy’s vision of using AI as an augmentation tool, one which extends our reach and ability to sense-make out of vast amounts of data.
But here’s the crucial catch: you must be the final checkpoint. AI models can sometimes “hallucinate,” meaning they might invent facts or sources. In fact, Sam Altman of OpenAI went on record recently saying, “People have a very high degree of trust in ChatGPT, which is interesting because AI hallucinates. It should be the tech that you don’t trust that much.”
The best approach is a tight “human-in-the-loop” system, something Karpathy argues for in his recent talk for Y Combinator‘s AI Startup School. Let the AI do the heavy lifting of gathering the data, but always ask it for its sources and then click those links to verify the information yourself. You provide the critical thinking; it provides the turbocharged research.
Understanding How a Large Language Model “Thinks”
To get the most out of these tools, it helps to peek under the hood and understand how they actually work. Karpathy seems to have the touch when it comes to articulating some pretty fundamental AI concepts in an elegant way – he’s the man behind the term “vibe-coding”. Another great ‘Karpathyism’ is his definition of an LLM as a “stochastic simulation of people.” That’s a fancy way of saying it’s a probability machine, which approximates and mimics the thoughts, ideas and beliefs of the most average or median human on Earth.
When you ask an AI to write a Stanford GSB essay on “What matters most to you, and why?”, it doesn’t feel or reason. Instead, it scans its memory of the entire internet—every essay example, every admissions blog, every forum post on the topic. Then, it starts predicting the most statistically likely word to start such an essay. “My,” perhaps? Or “I”? Or is the word “The”? Then, based on that word, it predicts the most likely second word, then the third, and so on. It’s essentially building the most average, most predictable, most “correct-sounding” essay it can, optimizing and finding the most probable or likely word, one at a time.
This is why purely AI-generated essays often feel so… bland. They lack that “zing” or unique edge because they are literally designed to be the median response. They are nothing but a collage of the most common themes and phrases ever used by humans in the context of a GSB application.
But you can turn this into a massive strategic advantage! Instead of asking the AI to write for you, ask it to show you the baseline. Prompt it with questions like:
- “Write me a typical ‘Why MBA?’ essay.”
- “What are the three most common career goals applicants write about?”
- “Give me a classic example of a ‘leadership’ story for an MBA essay.”
- “What’s the most common way applicants talk about how they plan to contribute and add value to their MBA peers and larger school community.”
The AI will hand you the most conventional, cliché-filled template. You might want to turn your nose up at such a draft, but this is gold! By seeing what the “most average” answer looks like, you now know exactly what to avoid. It gives you a clear benchmark to push against, ensuring your story is fresh, unexpected, and stands out from the crowd.
The Red Line: Why Your Story Needs a Human Heart
Now for the most important rule of all. While AI is fantastic for objective tasks, there is one area where you must draw a hard line: do not use it for qualitative feedback.
Asking an AI, “Is my story compelling?” or “Is my reason for wanting an MBA persuasive?” is a recipe for disaster. Why? Because any LLM is just a prediction machine, not a reasoning machine. It can’t feel empathy. It can’t understand the nuances of human ambition. It doesn’t know why the emotional arc of triumphing over a personal setback can capture the adcoms’ imagination.
When you ask for feedback, it simply compares your text to all the other texts it has seen (basically, everything ever written and uploaded to the internet) and gives you an answer based on the statistical average of everything. It’s an “argument from authority” where the authority is just a stochastic average of everything ever posted online. A measure of central tendency, as a statistician might like to call it.
Think about it this way: if the internet had existed in the 1500s, every single human textbook, blog, article, and reddit comment would have said the sun revolves around the Earth (well, everyone except Copernicus, I imagine). An AI trained on that data would tell you, with 100% confidence, that the geocentric model is a fact, and that the earth is indeed the centre of the universe. It can’t reason from first principles to arrive at a new, revolutionary truth like Copernicus did. It can only reflect the absolute consensus of its training data—flaws and all.
While AI is getting surprisingly good at logic and reasoning in rule-based systems or contexts like math or coding, your life story doesn’t have set rules. It has heart, nuance, and emotion. The very essence of a great application essay is its ability to connect with another human being. A token counting machine, by its very nature, cannot tell you if you’ve achieved that.
So, when it comes to the soul of your application—your stories, your “why”—you need a human heart to read it. Share your drafts with people you trust: mentors, alumni, current students, friends and family. They can provide the genuine, qualitative insight that a neural net transformer based on Bayesian probability simply cannot replicate, no matter how many billion parameters it can run, or how large its context window is.
So what’s the TL;DR?
Ultimately, navigating AI in your MBA applications is about balance. Embrace it as your tireless editor and brilliant research assistant. Use its predicting ability to sharpen your unique edge. But when it’s time to tell your story, trust in the one thing an AI will never have: your own irreplaceable human voice. That’s the story the admissions committee is waiting to hear.